Improving on "1-A"
Posted by Steve Paul
Video Player is loading.
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Resume Video
Start from Beginning
Watch Video
10
Loading 96 Comments...
Great first video!!!!! I like your thought on what is the best for training vids. i'm fully agree with you about theory/example vids rather than looking live/replayer stuff.
my question will be to what extent this concept can be applied to MTT?
even if you are semi-retired MTT player, I'd like your thought on how you can study to improve in MTT compared to cash game?
thanks again and looking forward to your next vids! great stuff, need the same in "essential MTT subscriber":)
Glad you liked the video! I think the concepts apply very well in the early stages of MTTs and even throughout. My feeling towards MTTs is that in general people try to make too many tournament specific adjustments. Most of the differences between cash and MTTs (except around the bubble and nearing final table where ICM considerations become much more important) come from the different stack sizes and the presence of antes. If you improve your cash play you will improve your early stage MTT play and likely the rest as well.
Most of the tournament specific improvements I made were putting in work with shove/call ranges with <10bb and 15-25bb stack resteal type situations. It's been years since I did any of that though.
Yeah, more theory videos!
Steve can you elaborate on the nuts/air situations and maximizing the EV of our whole range on multi streets (that we were discussing on your pgc), in some of the next installments.
Nice debut.
Yep, I'll put that on the list
Nice to see a focus on some math and theory in a lecture setting like this. It balances off the content on the site quite well.
Nice video. Was very informative and could easily be an elite video.
One suggestion I have: Also show the cards/players at the table. It would make it easier to visualize your examples if you show your examples at a poker table.
Also just want to clarify the concept you were explaining. Are you saying if button has a wide raising range from button(52%) we should be defending less than 58% of time? Using 1-A theory asserts that we should defend around 50% of time?
Also confused on how we know that someone is betting 100% of their nuts, or 30%. It seems good in theory, but when actually applied to the table you can't really determine the exact percents of time someone is bluffing, or value betting.
Interesting idea, I'll try to set something like that up for future videos.
I'm saying that on a K72r flop as bb facing a half pot cbet, 1-A would suggest we need to defend 67% to make his bluffs "breakeven". But that's not really useful, because button is debating between betting and checking, NOT betting and folding. Breakeven implies an EV=0. If we want his checking and betting options to have the same EV with his worst hands then we need to be defending something closer to 58%.
Alas this is the problem with theory, in practice the theoretically correct response is often not the max EV strategy because we're playing flawed opponents. But having an understanding of what approximately correct play looks like in a variety of situations is still useful.
"I'm saying that on a K72r flop as bb facing a half pot cbet, 1-A would suggest we need to defend 67% to make his bluffs "breakeven". But that's not really useful, because button is debating between betting and checking, NOT betting and folding. Breakeven implies an EV=0. If we want his checking and betting options to have the same EV with his worst hands then we need to be defending something closer to 58%."
Gotcha, so are you saying we should be calling with top 58% of hands in this instance? Once we call with top 58% of hands, is there a formula to figure out what type of boards we should be check/raising, check/calling, check/folding?
So using your example: the button raises top 52% of hands. We call with top 58% of hands. The math says this is the optimal play preflop. What does the math say about postflop play then?
I'm guessing we can apply 1-A to calculate hands we should be calling with in the BB? So if button is raising top 30% hands, we can use 1-A to figure out what % of hands would be optimal to call with?
Sorry I wasn't clear. I mean given the range we defended preflop, and assuming we don't check/raise the flop ever, we should check/call the flop with about 58% of the hands we get there with.
More generally 1-A can give you an estimate of how often you should defend against a bet/raise. In the video I'm trying to show some ways we can improve that estimate. We're still very very far from being able to say something like "it's optimal to call x% preflop vs a button open"
Thanks for the responses. Interesting to think about all of this. Should be applying more of this type of theory into my game. Been reading some poker books, but not ones that focus on the math like this. Going to start reading Janda, and Miller soon.
Do you have any tips for someone trying to get better at the mathematical side of the game? I think my math is above average for the midstakes, but still needs great improvement. Any good books that come to mind?
Highly recommend Will Tipton's two heads up books whether you play HU or not. Mathematics of Poker is a good read though not an easy one. I also enjoyed Janda's book.
Enjoyed this excellent video !
However, @09:30 you have an example where you have a range of 20 % trapping nut hands, 70 % bluffcatchers and 10 % pure air.
You say :
so we gonna call half of our hands, ( 20 nuts and 30 BC) and opponents BE on his bluffs :
- he loses 1 PSB when we call
- he wins 1 PSB when we foldBut that must be wrong, because we are not in a perfect PvBC situation.
We should either call with :
- 50 % of all our hands (your proposal)
- 50 % of all our hands that can beat a bluff (argued elsewhere)
-> thus 50 % X 90 % = 45 %
- 50 % of all our hands (except the trapping nuts) + all our trapping nuts
-> 20 % trapping nuts + 50% of 80% = 40 % = total of 60 %- 50 % of all our hands (except the trapping nuts) that can beat a bluff + all our trapping nuts
->20 % trapping nuts + 50% of 70% = 35 % = total of 55 %
Here, because we are not purely bluffcatching, our nut trapping hands will always call, and they don't call to bluffcatch, but to beat Villains value hands.
Also if we call with 30 of our 70 bluffcatchers, we only call 43 % of our bluffcatchers instead of 50 % so his bluffs are not indifferent to bluffing or checking (not taking into account the trapping hands calling).
When we call 55 % :
we make his bluffs indifferent to checking or betting with our BCwe make his value hands lose 20 % with our nut trapping hands
Anyway, this is a promising series ! I had not less than 2 aha moments, lol.
If we call 55% of the time (our nuts plus half our bluffcatchers) then he has 2 options.
1. Check and win the pot 10% of the time -> EV = 0.1
2. Bet. He'll get called and lose 1psb 55% of the time, get a fold and win the pot 45% of the time so
EV = .55*(-1) + .45(1) = -0.1
So his best play is to never bluff. Therefore call 55% cannot be our best response. Instead we call all of our nut hands and then enough bluffcatchers to make his bluffs indifferent. In this case that's 25 bluffcatchers so we call 45% of the time (as shown in the video). So we actually only call 36% of our bluffcatchers here.
edit: also, though the idea of "call half our range that beats a bluff" gives the correct answer of calling 45% of the time in this case, I don't recommend using it as it doesn't always work. For anyone interested why, use the same river range but have us be in position on the river (ie we can bet if villain checks.)
You also state that we have trapping nut hands, so if we're trapping, it's
In the case Villain doesn't know our range, the correct frequency is one ofonly if Villain doesn't know our range, right ?
the 4 options.
1. Check and win the pot 10% of the time -> EV = 0.1
that seems obviously impossible : if he's reasonably polarized he shouldbeat most if not all our bluffcatchers with his value hands on top of the 10 %
air we have; his EV is automatically > 0.1
I'm a bit confused here, having a few traps and villain knowing we have some traps is fine. Villain is still correct to value bet all his hands better than our bluffcatchers as he wins vs 25 of our calling hands and loses to 20. If we had many more traps then this scenario wouldn't really work.
Yes you're right I should have been more clear. The EV of checking his bluffs is 0.1 and these are the hands I'm focused on when designing my river calling range. The EV of his value hands is 0.8 if he checks (he beats all but our nuts) and 0.85 if he jams (.55*1+.25*2+.2*(-1)).
Also, could you please post a reference to the hands you mentioned posted in the forums ? tx.
dodgybob post is here: http://www.runitonce.com/nlhe/designing-flop-ranges-btn-v-bb-k72r/
thank you.
And yeah, it would be even more useful if you could illustrate the theoretic example (like the above mentioned) with 1 or 2 realistic example ranges and board.
Welcome to the team Steve! Funny my next vid is also on the dodgybob post.
Very happy to see some good theory content for essential members. Looking forward to your future videos.
Welcome! Awesome first video, Steve.
Nice vid. Realy like the pace!
Steve, great first video. I thought the content was very good, and your presentation was nearly perfect.
you have been talking about exactly the things that i have been thinking about recently and cleared alot of stuff up for me, very happy about this video nice job.
One more question about defending in the bb vs cbets of our opponent: Is there some way to determine for individual hands how good they play vs our opponents assumed cbetting range, when looking at scenarios where we are contemplating check/calling? I mean for example when we face a cbet of 100 into 200, if we called and we would go to showdown immediately right after with no further action, we would need 25% equity. However, of course our opponent will barrell us off our equity some of the time on turns and rivers, so i would say we would in general need more equity vs opponents range on flop than 25% to be able to make a good long term call.
I know there is no exact estimation that you can simply use all the time (like i need X% more equity than the showdown equity i have on flop to be able to make a call vs opponents range), since it depends on opponents aggression factor and tendencies plus on flop structure + remaining stacksizes etc, but is there maybe a range of flop showdown equity in which we can operate, depending on the named factors?
I am an MTT player so im still looking mainly for exploitative plays, im interested in balanced plays too though since as you said it is always a good starting point to know. But also, does the fact that our opponent barrells on turns and rivers not effect our balanced flop defending frequency? I dont know how it will look on turns but if we get to turns on average with our range consisting of weak hands to a large degree, wont we have to defend too much on turns and therefore also on rivers if we always simply go by minimum defence freuqencies starting from the flop onwards? Especially when ranges are wide to start with (sb vs bb or button vs bb), it seems to me like we will have a hard time finding a reasonable calling range by the river if we go ccall (minus sraises) min defence flop, call (-raises) min defence turn, call (-raises) min defence river.
Maybe we can become hard to extract value from, but also would that even be realistic to our opponents behaviour or wouldnt we lose a ton of value not making exploitative plays as realistically our opponents will never be that aggressive? You named that extreme example of someone having 100% nuts, here an extreme example would be a player never bluffing only always continueing with value. I would argue players just dont continue enough in most spots in reality that we should not rather tighten up a bit.
Sigh already a wall of text but also since you come from an MTT background as well, in MTTs our calling frequencies should tighten up even a bit more as ICM is a factor and stack utility risk etc so taking very marginal edges is often not good right?
A lot to potentially talk about, I'll give a couple thoughts. First, you're right that some hands aren't going to do as well as their equity would indicate. But others will do better! This is almost certainly something you already intuitively know - some hands (eg draws) play a lot better than their equity would indicate and others (weak pairs with little chance to improve) play a lot worse. So you can't just multiply your equity by some number and decide if you can defend, it's going to depend on how your hand plays on future streets. The more someone barrels the less valuable your weak made hands are (up to a point where they're just way overbluffing) and the more valuable your hands that can improve.
As for MTT specific adjustments, I'd avoid making huge adjustments from what you believe is correct except around bubbles and when exploiting weaker players. Also keep in mind that mtt regs (in my experience) bet a lot smaller so you just need to call wider unless you believe they're not barreling enough. If you have a read you can exploit, go right ahead and exploit it. I'd take my advice with a big grain of salt, it's been a long time since I played MTTs.
Really appreciate all the positive feedback, thanks!
Would love more on this theme. For example river spots where we should give opponent profitable bluffs due to the action on previous streets, three-street scenarios where we are at a range disadvantage, scenarios where it hurts us too muh to make his bluffs indifferent to checking/betting because it gives away too much value to his stronger range or because it leads to us folding too many profitable hands, etc.
@14:45, you state :
to make Villain indifferent to bluffing vs checking his bluffs you want to call x% of your [ whole ] range, not x% of your bluffcatchers
imo, that is what is stated in MoP, however it's sometimes argued the opposite :
GTO simplified (OTR)what do you think ?
Hi,
I thought this was a really interesting first video. I like ones focused more on theory rather than live sweats and I thought this one struck a good balance between getting the important points across without being too dry or math intensive.
I don't know what your future plans are or if this is going to be a series, but I would really like to see some hands worked out start to finish. I understand you start at river situations because they're the easiest to visualize and apply these theories to if you have a range of pure air and pure nuts, and then I understand why you go to a common flop situation that we all get into all the time. What I'd really like to see is a continuation of this flop example, so we call with our range on the K72r board and get some kind of turn and face another bet. At this point we'll probably want to check raise with some sets, two pairs, and maybe some back door straight and flush draws. How do we construct that range, as well as our range for calling? Then we find ourselves at the river after check/calling twice. What do we need to call with the third time? Can we have a river check/raising range that includes bluffs? Then maybe do it all again with a wetter board :) I think this would really help me from playing every street in a vacuum.
hi Steve,
At around 24:45 you say that for specific spots (i.e. checking back air OTF) you only need a relatively small sample to get a decent idea of the true EV of those spots. Could you explain why?
they're going to be relatively low variance spots since I'm ~never playing a big pot. As I think more about it I'm not sure I'm right other than that though...on my laptop right now but I'll try to remember to check my db for the standard deviation in those hands next time I'm on my desktop
thanks for the responses =)
Link to pgc thread please steve
http://www.runitonce.com/chatter/200-zoom-and-up/#123808
It's in the RIO forums under chatter->poker journals
I agree with the others regarding this video. Excellent, well presented and easy to understand.
Great video!
Just wanted to jump in and say well done.
really good video. when our river chk range is 70bc/20n/10a and villain is deciding whether to check or bluff, you assumed that his possible bluffs beat our air. It seems like OOP air would at least be flipping with in position's air in most scenarios. So I'm not sure so much EV is often available in a checkback as in the video examples.
Agreed that the 70/20/10 example is not realistic at all, really just the general conclusion I was interested in there
Hey great video man! Watched it twice.
Just had a few questions about some of the concepts.
1) In the nuts/air vs bluffcatcher example. You state that that villain has 30% nuts and 70% air on the river with 1psb. Since he has 1psb, in order to meet 1-a, we must defend 50% of the time. However, we lose to his value hands, no? So we should actually defend less than 1-a since his range is protected with his nut combos. So we only defend 1-a if his range is composed entirely of bluffs? I understand these examples are simplifications, so I'm only asking about this because you explicitly state that villain has 30% nuts. I'm assuming there are modifications to 1-a that can be made to account for the nut combos in villains polarized range. Is this an example of where we would need to make a modification to account for his 30% nut range?
2) In the K72 BB vs Btn example, are we making the simplification that btn's entire range is composed of the hands from the bottom of his range with 13-17% equity? If btn cbets his entire range on K72, his EV vs our continuing range is much higher than just 1BB. I realize that 1-a is a computation for how often we should defend vs his BLUFFS, but since villains range is not composed entirely of his minimum equity bluffs when he cbets K72, are we just ignoring the rest of his range? Similarly, are we assuming that all of his check back range is his min equity hands. Since if he checks back his entire range (which is what were trying to make him indifferent between, right?), his entire range should have much more equity than the 1.3BB?
Am I just overlooking some simplifications that are being made? The logic seems completely sound in situations where villains range is all bluffs (or his min equity hands that we made the adjustments for). But, when you consider his entire range, like his value combos etc., the math doesn't seem to account for them. Or is this another modification to 1-a that you will address in a future video? Or am I just making a logical thought process flaw here?
Anyways, once again, awesome video! I really enjoyed it. Hope to see more from you in the future!
1: If we call less than 50% of the time, his correct adjustment is to bluff all of his air. We want to prevent that from being a profitable deision. His value hands getting a bet 50% of the time is unfortunate but necessary to be unexploitable. If we play differently (ie exploitively) we may do better vs his particular strategy but we do worse vs his best response.
2: We can't make him indifferent between betting/checking his entire range. Realistically we can make some hand (or small group of hands) indifferent, with a bunch of hands being profitable (semi)bluffs and a bunch being better checkbacks. What this target hand should be I have no idea. In the video I chose his very worst hands which is ~certainly not correct but allowed me to illustrate the idea/process.
Very very generally speaking strong value hands are hard to make, it's ok for them to make a lot of money. Crappy hands are really easy to make, it's not ok for them to make a lot of money.
I agree with the others, you combined a good subject matter with such a nice tempo. I guess there's always that balance where you want to be clear to those unfamiliar with the subject without boring the ones that are, and you struck a perfect middle ground :) Looking forward to more!
This is definitely a balance I've struggled with in teaching and I'm sure I will have issues with it in future videos. Got some first video rungood that people find this one is well paced :)
Nice video, worked on such topic myself recently, w combining all typical values to excel sheets. There is lots videos w live sessions, so video on theory concepts w apllication examples are really nice. Maybe would be good to do some videos w analizying field populations of different limits in general, to find some tendecies that could be exploitable.
If oop construct his range such that J5hh/JTo is slightly +ev both betting and checking on Ks7d2c, which range should they go in?
How much do you think button should defend vs 3/4th pot when checking back the flop or turn?
Can either of buttons ranges be 'unbalanced' in GTO play, ie. folding a lot more than the odds offered if bet or raised against ?
How do you deal with the fact that every 'value' hand button have blocks made hands and every bluff doesn't and oop will similarly block value-hands with his bluff-catching range.
Bet or check KK as BT?
If the EVs are the same then in theory they should probably be split between both (not necessarily equally). In practice, whichever one you want.
Haven't done any math but depends a lot on his cbetting strategy. For the typical high cbet strats less than 1-a is totally fine.
I'm not sure what exactly you mean here and we're getting way beyond the scope of the video. You can't fold to a raise such that raise>fold for BB though.
I don't know :) Interesting result in the 25bb preflop only minraise/jam or fold game is that bb jams less than 50% of all hands but button has some open folds (because card removal pushes bb's jam range past 50%). So card removal matters, but we already knew that.
I check but can't prove it's right.
Wow. Only watched first minute so far but this is creepy.
Started playing 2004. Check
Got married 2008: Check
Started Family in 2009: Check
Teacher certificate in Chem and Math: Check
Going to keep on watching now but if this continues I'm going to start wondering if I made this video and I'm getting dementia.
Rly good video Steve! You're very good at teaching, and def know your math. Powerpoint was flawless imo, I think it's quite weird that this is a essential video. Steve should be elite imo.
Alright finished watching now. Great first video!
Really looking forward to future videos when you might start to get into some more advanced topics.
we should have more videos like this is essential videos
Yeah, I watched this and didn't even know it was an essential video. I just assumed it was elite because of the content. I think this will be huge for essential members and think its very good that RIO has added another very knowledgeable coach to its essential roster.
Ha, Im also an mtt player who moved to cash (just recently) and who studies to become a teacher. Great video, looking forward to moar!
Nice Video! My only comment would be that when defending oop, we're not trying to make the opponent indifferent between betting and checking his worse hand (in your K72r example, the hand you used was something like 54 no back door), but instead the likely indifferent hand for the opponent will be a better bluffing hand which will have MORE EQUITY than the worse hand.
Agreed, definitely something I should have mentioned! I didn't want to get too much into picking a hand (or small set of hands) to make indifferent but watching the video you could easily get the wrong impression that it's only about the worst hand.
How do you estimate which bluffing hand needs to be indifferent?
Unfortunately I have no good answer to that question.
Hey Steve Great video
Whats the name of the software you use in the video?
Thanks! Software is from http://www.cardrunnersev.com/
Excellent video. Really enjoyed it.
Very good video! I was looking for something like this for ages...
One big question though:
In most situations preflop ranges are asymetric. Doesn't that have an impact on the minimum defence frequency?
I mean in most situations opponent's range is narrower/looser than ours, our range or our opponent's is capped, etc. Shouldn't we take that into consideration? How?
Thanks so much for this video, very cool.
grunch...sorry if this was discussed already:
In the 20/70/10 nuts/bluffcatcher/air example, why did you assume that villain's bluffs would win against our air at showdown if he checks back?
Also, and sorry if I'm just nitpicking an example you were using to make a point, but in this situation (if our air is guaranteed to lose at showdown) wouldn't it always be correct for us to lead the 20 nut combos and the 10 air combos or something like that (and then call with just the right amount of bluffcatchers)?
Both your concerns are valid - was just using a contrived example I'd seen in the forums to illustrate the point. Glad you liked the video!
Steve, what's wrong with this calc?
If BB fold 42% of the time, the calc will be:
(0.42 * 5.5) + (0.58 * -2.75) = 0.715bb
+ the 1bb from our 5d3d equity
= +1.715bb by c-betting.
But checking is 1.3bb.
Shouldn't BTN always c-bet then?
I made this video ~8 months ago, so the details aren't exactly fresh in my mind...but it appears your mistake is that you're adding 1bb 100% of the time, we don't get that when he folds
Thank you very much for teaching the 1-A. I'm happy to be encountered in this school. I'm looking forward to see more from your videos, great job.
Simon
Thanks!
Thanks for the video, Steve!
I got a little bit lost in one simple spot. Around 23.55 from the start of the video you have mentioned, that if we are open folding the flop with low equity "give-up" hands after 2.5x open raising preflop, we are playing with -250bb/100hs winrate. Please, correct me, if I'm wrong, I've always thought that EV of folding equals zero.
Looking at it on the flop, the EV of folding is 0. But for the whole hand we've lost 2.5bb, or 250bb/100hands.
Hi Steve,
Great video.
One question - my understanding was that when we defend the BB vs a BTN minraise, for example, we are getting such good odds that we defend very wide, but this preflop decision leads to a situation where we have a very weak range on the flop, and so will be giving up "more than we should", ie more than we should if we had chosen to put 2bbs into the pot instead of just 1, if you see what I mean.
Sorry if this is a little vague but I suppose my question is can it be correct for us to "over fold" flops since we got such a big discount preflop, and we therefore "can" allow villain a profitable bluff on the flop, since in order to get to this profitable situation he had to risk money by raising preflop.
Hopefully you see what I mean, I know this is an old video but I really enjoyed it and am planning on watching the rest of your theory videos so any comments would be much appreciated!
That's kind of what this video is about. The default response to how often should I defend vs a cbet is "1-a" but here I show some reasons why we should generally defend less than that. Having a weak range on the flop means that his EV of checking even his weakest hands is going to be >> 0, so we shouldn't aim to make the EV of his cbets 0. ie we should defend less than 1-a
Thanks for the reply.
Yeah I figured as much, its just that the discount for seeing the flop wasn't specifically mentioned, I suppose I wanted to check if I was making a mistakes by allowing the previous street's action to influence my defence frequency on the following street.
Hi Steve --
Thank you so much for the video -- it was very helpful in breaking down some concepts for me. I am, however, getting stuck on a very basic concept & was hoping you could clarify, even though the video is old:
In an example you give early on (the 20% nuts/70% BC/10% air one), you initially calculate the defense frequency needed to make EV(bluff) for V = 0. You then note that V can instead check back, and EV(check) = 0.1 (assuming V beats your air). You then say that we should modify the defense frequency such that EV(bluff) also = 0.1.
I understand that this is a key concept for game theoretic approaches, but I don't understand why it's true; I suspect the problem is in how I'm coming at the issue. I approach it as:
* If we select a defense frequency such that EV(bluff) = 0 and EV(check) = 0.1, V will sometimes bluff and sometimes check, and the average EV of his actions will be between 0 and 0.1. If V plays perfectly, he will always check back given these options, in which case his equity across all actions will be 0.1. If V plays less than perfectly, however, he will occasionally bluff, and his equity across all actions will be slightly less.
* If we select a defense frequency such that EV(bluff) = EV(check) = 0.1, no matter what V does, his EV is 0.1.
From the way I'm approaching this, this makes it seem like choosing a defense frequency such that V is indifferent between checking and bluffing basically gives him an EV freeroll, and is a strictly worse strategy than choosing a defense frequency such that EV(bluff)=0 (i.e. can yield the same or better, but never worse, results).
Can you set me straight?
to make his bluffs EV 0 you call his bets more often, increasing the EV of his value hands. It's not just about minimizing the EV of his bluffing region. To prove it to yourself, find the EV of his whole range against a couple different strategies and see what the results are.
Let me just close the door of my DeLorean.
You say we should make some EV- calls to make him inddifferent from betting and checking. Because if he's better checking he wouldn't bluff right?
So what's the problem if he doesn't bluff? or bluff enough?
We shouldn't make any -EV calls. If you are trying to play optimally, your goal is to minimize the EV of an optimal playing opponent. Generally if an opponent deviates from optimal play he will lose money (or at least not gain). If you play in a way that his best play is to never bluff and he never bluffs then you just lose too much to his value hands with nothing to make up for that.
But if we know he never bluffs, we only be calling with our nut hands right? Because he doesn't have enough bluffs.
but then he could always bluff? Not sure I'm understanding your question
You did understand, your response answer my question, so we could still adapt to his play by calling, and then he would adapt to ours.
But what I'm getting at is that his play isn't optimal to begin with, or at least with that size. His optimal play should be a bet with a size X wich is a function (in other words, it would depend on) the eq of his range against ours.
The situation where he can bet another size might be outside the scope of the toy game, but for a real case might have worth analyze it.
EDIT: Maybe I misinterpreted you in the video. So I'll make sure I didn't.
All the bluff catching calls that we do are in worst case scenario EV=0 right?
If answer is yes I should pay more attention to the video and probably eat healthier.
yeah an optimal strategy does not make any -EV calls vs an optimal opponent.
Hi Steve, I still don't really get how alternate lines can affect calling frequency. Why does how often we call depend on how often we check the river? Is there a correlation such as: the more/less we check, the more/less we'll have to call?
Here, you emphasize the importance of "what are we trying to accomplish?" And, as you stated, we want to call enough of his river bets such that he cannot profitably bluff the river with any two cards from his range. In the extreme example you gave where we are betting the river 99% of the time, does this very high betting frequency (and thus very low checking frequency) help accomplish our objective because it is denying him the opportunity of bluffing on the river since we are leading out first?
Lastly, just to better understand this concept, can you provide other example alternate lines we could take that would also accomplish our objective?
Thanks
The less often we check, the less EV he gets from being able to bluff profitably. If we are very frequently checking then it is more of a concern. Your river strategy's goal is to minimize his overall EV. So in the extreme case where you get to river with nuts/air, you will of course never call a river bet even though you will be checking (to give up) some reasonable % of the time. This also shows that what we're trying to accomplish isn't really to not let him bluff any two cards, but just to minimize his overall EV.
Steve - late to this video but a quick question. You explain the concept that when IP check-behind has EV>0, then we don't need to defend 1-a because he SHOULD have EV>0 when he bets in order to be indifferent. My question is: if we reverse this so that Villain is OOP, his only opens are to bet or check. Is it more likely that 1-a will be correct in that scenario (excluding an analysis of the EV of a Villain check-raise)?
it will probably be closer but checking as OOP still has EV>=0
but its not just what he bets at 4 min of video. Its also how often he bets and if its a lot adding more bluffs in his range then we have to call more based on his repetition of doing it at that betting amount. Your saying math wise we need to call 66% of the time but if hes only bluffing 10% of the time then we are losing money. i think figuring out how many times hes bluffing with that bet size is more important then anything. then we can figure out how often we need to call where he cant exploit us.
You're talking about maximizing your EV vs a weak opponent. At 4 minutes we're talking about calling such that we can't be exploited by an opponent who is over/under bluffing.
Whether an opponent is weak or strong is irrelevant it only matters what their tendencies are and using the proper ranges vs what they are doing so we cant be exploited.
If you are adapting to your opponent's tendencies then by definition you are exploitable (they are allowed to adapt/adjust too!)
Great stuff
Hi steve, Nice video!
I am having a really hard time understanding the concept of MDF, It's basically the defense frequency for which the opponent is indifferent between betting and checking right? And it's an optimal defending frequency right?
Now let's say we start defending less than MDF. Agreed that his bluffs will gain EV, but wouldn't he lose EV with his valuebets as we are defending less than before. And isn't it possible that ev loss exceeds the EV gain from bluffing.
For eg, let's say villain range is 80% value, 20% bluffs:
Now let's say we call by MDF, now EV(villain)= 2/3pot+ 1/3rd(4/5(3pot)-1/5(2pot))
EV(villain)=4/3*pot
Now say we adjust by now defending at all.
EV(villain)=pot
Now how is MDF the most optimal strategy?
I might be thinking everything wrong. Can you please clear my misconception.
You can see more about this in my 2 part series on betsizing but very often MDF is not the same as optimal defending frequency. In your particular example you should not defend MDF and if your range is all bluffcatchers you should fold everything no matter how small he bets. Again, you can see my betsizing videos (and the comments sections of those videos) for a more in depth explanation.
Really enjoyed this video. Thanks Steve.
Steve, on rewatching this video, another doubt popped up in my head.
It is are we supposed to defend 1-alpha of the hands which can beat a bluffcatcher, or is it 1-alpha of our range ? Or now since his bluffs have some checking ev, we will like to make him indifferent between betting and checking ? In that case what if he bets small enough that we are supposed to defend some of our air hands, he will gain Ev right?
1-a works most cleanly in nuts/air vs bluffcatcher type situations. For the broader case, making some bluffing hand (or set of hands) indifferent between betting and checking is the better way of thinking about it - though there are situations where indifference breaks down as well (see my betsizing videos for an example)
Steve great video! Question: if I bet turn, get called and get to a river spot oop where I'll only check 40% of times, and the rest I'll jam, there's a PSB left....can I get away with always check/folding BC's to a jam since villain's floats only get that opportunity (of me checking river) less than mdf? I'm asking because I don't know if that's what's implied on the example you gave at the 16 minute mark....
"It doesn't matter if I always check/fold if I'm only checking 1%". That's what made me think of this question. If that statement of yours is strictly true (and I think it is), then I guess I'm on the right track here?
Be the first to add a comment
You must upgrade your account to leave a comment.