100NL Coolers or bad play?
Should be clear that it's for value, no? Obv never folding out better.
Yeah, but you're blowing up the pot with a hand that doesn't want a blown up pot right?
Well, very often you're simply ending the hand and claiming the pot, since villain should be folding to a c/r ~30-40% of the time, and this is a significant victory. The only hands that WANT to blow up the pot (i.e. get called or 3-bet after a c/r) are strong 8x really, but obviously we want to be c/r wider than that. The question is which non-8 hands are best to c/r (and which non-8 hands are best to c/c). Although 4x may seem/feel like it should be a better "bluff-catcher" (hand to c/c flop with) than K7+bdfd, and K7+bdfd the better "bluff"/"semi-bluff," I'm pretty sure the reverse is true.
Well, very often you're simply ending the hand and claiming the pot, since villain should be folding to a c/r ~30-40% of the time, and this is a significant victory. The only hands that WANT to blow up the pot (i.e. get called or 3-bet after a c/r) are strong 8x really, but obviously we want to be c/r wider than that.
The question is which non-8 hands are best to c/r (and which non-8 hands are best to c/c). Although 4x may seem/feel like it should be a better "bluff-catcher" (hand to c/c flop with) than K7+bdfd, and K7+bdfd the better "bluff"/"semi-bluff," I'm pretty sure the reverse is true.
There are other hands you can checkraise, like gutshots, 96ss or T9?
On the river, it is absolutely correct that we want to polarize our betting/raising ranges, so raising a pair of 4s and calling with king-high would almost never make sense (only if blockers were relevant, basically). However, hand values on the flop aren't static, so it can sometimes be (I would even argue that it usually is) correct to have multiple discontinuities in our betting/raising ranges. Thus, when viewing a list of our hands in order of their absolute strength on the flop, we may want to raise our strongest hands, call some hands weaker than that, raise some hands weaker than that, fold some hands weaker than that, call some hands weaker than that, raise some hands weaker than that, and fold some hands weaker than that (or almost any other permutation of those options, really).
The fact that so many people c/c the flop here with 4x is probably a large part of the reason why it seems like this board offers a large ex-showdown equity advantage to the in-position player. If you have a flop c/c range that is too light in 8x hands, you WILL get owned on later streets. The same is true of your c/r range, of course, but it's my claim that your c/c and c/r ranges should have very similar frequencies of 8x in them (20-30%, basically). Where I'm pretty sure my understanding of this board differs from most, though, is that I think you should be more likely to hold an 8x hand after c/c than after c/r (but also that you should be much more likely to c/r an 8x hand than c/c with it).
If your non-8 c/r hands will lack showdown value on the turn/river too frequently, then you become susceptible to frequent small flop 3-bets and/or floats with weak hands. Hence it seems best to me to usually c/r your strongest (and weakest! Check the board - see why?) 4x hands for value rather than c/c.
Those are all good hands to check-raise (assuming you have a backdoor flush draw with the T9/96), but I think you should be check-raising much more than that (and I'd almost always 3-bet T9s, so it wouldn't be available to me for check-raising here).
If you're x/r-ing a 1:2 ratio of "value":"bluff" hands on the flop (with "value" defined here as 8x), you should be able to comfortably contend with floats and/or small 3-bets even if you're hardly ever x/r-ing 4x.
I've discussed elsewhere itt why I don't think using terms like "air" (or in this case "airy") is helpful and can in fact encourage sloppy thinking/poor play.
On the river, it is absolutely correct that we want to polarize our betting/raising ranges, so raising a pair of 4s and calling with king-high would almost never make sense (only if blockers were relevant, basically). However, hand values on the flop aren't static, so it can sometimes be (I would even argue that it usually is) correct to have multiple discontinuities in our betting/raising ranges. Thus, when viewing a list of our hands in order of their absolute strength on the flop, we may want to raise our strongest hands, call some hands weaker than that, raise some hands weaker than that, fold some hands weaker than that, call some hands weaker than that, raise some hands weaker than that, and fold some hands weaker than that (or almost any other permutation of those options, really).
On the river, it is absolutely correct that we want to polarize our betting/raising ranges, so raising a pair of 4s and calling with king-high would almost never make sense (only if blockers were relevant, basically). However, hand values on the flop aren't static, so it can sometimes be (I would even argue that it usually is) correct to have multiple discontinuities in our betting/raising ranges. Thus, when viewing a list of our hands in order of their absolute strength on the flop, we may want to raise our strongest hands, call some hands weaker than that, raise some hands weaker than that, fold some hands weaker than that, call some hands weaker than that, raise some hands weaker than that, and fold some hands weaker than that (or almost any other permutation of those options, really).
The point of check-raising 4x is not to help our 8x get paid off, it's to play the hand in the most profitable possible manner.
The fact that so many people c/c the flop here with 4x is probably a large part of the reason why it seems like this board offers a large ex-showdown equity advantage to the in-position player. If you have a flop c/c range that is too light in 8x hands, you WILL get owned on later streets. The same is true of your c/r range, of course, but it's my claim that your c/c and c/r ranges should have very similar frequencies of 8x in them (20-30%, basically).
The fact that so many people c/c the flop here with 4x is probably a large part of the reason why it seems like this board offers a large ex-showdown equity advantage to the in-position player. If you have a flop c/c range that is too light in 8x hands, you WILL get owned on later streets. The same is true of your c/r range, of course, but it's my claim that your c/c and c/r ranges should have very similar frequencies of 8x in them (20-30%, basically).
Where I'm pretty sure my understanding of this board differs from most, though, is that I think you should be more likely to hold an 8x hand after c/c than after c/r (but also that you should be much more likely to c/r an 8x hand than c/c with it).
much more likely to c/r an 8x than to c/c it:
10% 8x total, 60% of that goes into checkraise range for 6% checkraise. 1:4 ratio, so total checkraise = 30%
that leaves us with 40% 8x (4% of total range) to check/call with
more likely to have an eight when c/c than when checkraising means 8x should be >20% of a c/c range, so total checkcallrange should be <20%
If you accurately described your flop strategy here this means your flop strat looks roughly as follows:
30% raise
19% call
51% fold
Correct?
If your non-8 c/r hands will lack showdown value on the turn/river too frequently, then you become susceptible to frequent small flop 3-bets and/or floats with weak hands. Hence it seems best to me to usually c/r your strongest (and weakest! Check the board - see why?) 4x hands for value rather than c/c.
Can't really think of a good reason why you'd checkraise your weakest 4x over middle 4x. Ofcourse sometimes the board will come say QQ and your 54 will be worthless, but idk if that compensates for the times where it goes c/r checkdown and you lose vs J4.
So in practice:
much more likely to c/r an 8x than to c/c it:
10% 8x total, 60% of that goes into checkraise range for 6% checkraise. 1:4 ratio, so total checkraise = 30%
that leaves us with 40% 8x (4% of total range) to check/call with
more likely to have an eight when c/c than when checkraising means 8x should be >20% of a c/c range, so total checkcallrange should be <20%
If you accurately described your flop strategy here this means your flop strat looks roughly as follows:
30% raise
19% call
51% fold
Correct?
much more likely to c/r an 8x than to c/c it:
10% 8x total, 60% of that goes into checkraise range for 6% checkraise. 1:4 ratio, so total checkraise = 30%
that leaves us with 40% 8x (4% of total range) to check/call with
more likely to have an eight when c/c than when checkraising means 8x should be >20% of a c/c range, so total checkcallrange should be <20%
If you accurately described your flop strategy here this means your flop strat looks roughly as follows:
30% raise
19% call
51% fold
Correct?
After having watched you reverse engineer my flop strategy from a few casual comments, I am now moderately concerned about giving away too much free info. =P The possibility of a higher running pair was part of the reason, another was the possibility of turning a pair+gutshot, another was the value of hitting your kicker after c/c vs. after c/r.
If your flop folding frequency is 51% thats exploitable. I thought your strategy was designed to have the highest EV possibile while remaining unexploitable.
Against a 70% flop cbet, given he still has equity with his bluffs shouldnt we be folding < 40% of the time.
Depends on bet size. Against a 70% pot c-bet, yes, that sounds about right.
How d'you figure?
Well, that's impossible. Unless your opponent is playing the Nash equilibrium, the highest-EV strategy possible will always be exploitable.
My in-game strategy is to approximate Nash as a default and to deviate from my default whenever I have a pretty good idea that doing so will increase my EV.
No. Why would you think that? It seems like you're assuming that the SB shouldn't be able to profitably bet any two cards, but it's not clear why that'd be a good assumption to make.
My in-game strategy is to approximate Nash as a default and to deviate from my default whenever I have a pretty good idea that doing so will increase my EV.
No. Why would you think that? It seems like you're assuming that the SB shouldn't be able to profitably bet any two cards, but it's not clear why that'd be a good assumption to make.
Because against a 3/4 size pot bet we "should be folding" < 40% of the time otherwise any 2 cards can profitable bluff. Shouldnt the theory be that we want this number as close as possible so that any 2 card cant profitably bluff.
My in-game strategy is to approximate Nash as a default and to deviate from my default whenever I have a pretty good idea that doing so will increase my EV.
No. Why would you think that? It seems like you're assuming that the SB shouldn't be able to profitably bet any two cards, but it's not clear why that'd be a good assumption to make.
If thats the case are you betting with any 2 cards here ( I mean air - yeah i know you dont like to use air but sake of simplicity say anything less than Khi but the 2 cards are greater than the 4), and what implication does that have on latter streets against a person folding <40%?
FWIW I fold < 40% on this flop (even less against smaller bets), and often don't c/r this flop either. I just have a much stronger c/c range. I dont see why thats so bad....
Why?
You mean the pre-flop raiser? He hasn't c-bet yet at the point where I offered that evaluation. The fact that betting should get more folds on 884r than on J87s is not in conflict with the claim that the ex-showdown equity advantage of the in-position player is larger on the latter board than on the former.
Post your c/c range.
This is a judgement call though right? I feel like folding 51% is too much, can't really quantify it well but I'm assuming you can't either.
Disagree on the second statement, but that agrees with your point of folding flop more.
No; scroll up and re-read the thread.
Sloppy wording! They wouldn't make unprofitable river calls, they'd just make light ones, and that leads to the BTN being able to valuebet lighter. But I don't think we have any idea what the bottom GTO valuebet should be. I'm assuming you'd say GTO is to only go for 3 streets of value with 8x+, but maybe GTO is that overpairs+ should be able to go for 3 streets of value...
The fact that betting should get more folds on 884r than on J87s is not in conflict with the claim that the ex-showdown equity advantage of the in-position player is larger on the latter board than on the former.
No; scroll up and re-read the thread.
The person folding <40% on the flop will be forced to fold turns and rivers so frequently that they'd have been better off just folding the flop. Alternatively, they can continue to make unprofitable calls on the turn/river and thereby donate even more.
Right, but it's one I'm extremely confident in.
I wouldn't be very surprised if 51% was too much, but I would be EXTREMELY surprised if GTO play did not involve folding enough that betting ATC would be profitable for the SB. To put it concretely, I'd take 1.5:1 on either side of a bet over whether the GTO continuing range was > or < 49%, but I would lay 50:1 that the GTO continuing range is narrow enough that betting ATC would be profitable for the SB.
You're assuming that equilibrium bluffing/calling river strategies do not make use of card removal effects, which I'm not sure is correct. If they do, then many (not all) light river calls would be unprofitable.
It depends on how much you bet and how the board runs out, obviously. I wouldn't be at all surprised if there were some reasonable sizings/run-outs where getting 3 streets of value was possible at equilibrium. That doesn't necessarily mean betting will have a higher EV than checking at some point though.
It depends on how much you bet and how the board runs out, obviously. I wouldn't be at all surprised if there were some reasonable sizings/run-outs where getting 3 streets of value was possible at equilibrium. That doesn't necessarily mean betting will have a higher EV than checking at some point though.
I took a look at checkraising 5x on 995r and I wasn't particularly convinced that A5 is a better c/r than c/c (or even close). I also don't really see why a strategy that only calls flop would do worse than your one.
If only there was a way to take you up on this. Not that I think you're wrong, but I'd lay like 10:1 or something which is a huge difference
I don't assume they don't, because they do.
Say you have two type of bluffcatchers, with and without card removal effects ("with" meaning beneficial for bluffcatching). If your range is tighter and you only need to call the bluffcatchers with card removal vs a given betsize, equilibrium strategy for BTN is to make those bluffcatchers indifferent. If your range is wider and you need to call the bluffcatchers without card removal, equilibrium strategy for BTN is to make those indifferent, and the good bluffcatchers just become slightly profitable calls.
There can be no such thing as an unprofitable river call when playing vs GTO, because you could just exploit it by folding
*edit* in theory it should be possible that every single bluffcatcher you call with is profitable except for the very bottom combination, which would be indifferent
Yeah ok, the point is that we both assume "for this runout and this sizing, hand xx is the bottom hand that goes bet/bet/bet in GTO". What we assume this hand to be decides our flop calling (and thus also continueing) range. I don't have a solid method to decide what hand XX is, do you?
But I would lay 50:1 that the GTO continuing range is narrow enough that betting ATC would be profitable for the SB.
You're assuming that equilibrium bluffing/calling river strategies do not make use of card removal effects, which I'm not sure is correct. If they do, then many (not all) light river calls would be unprofitable.
Say you have two type of bluffcatchers, with and without card removal effects ("with" meaning beneficial for bluffcatching). If your range is tighter and you only need to call the bluffcatchers with card removal vs a given betsize, equilibrium strategy for BTN is to make those bluffcatchers indifferent. If your range is wider and you need to call the bluffcatchers without card removal, equilibrium strategy for BTN is to make those indifferent, and the good bluffcatchers just become slightly profitable calls.
There can be no such thing as an unprofitable river call when playing vs GTO, because you could just exploit it by folding
*edit* in theory it should be possible that every single bluffcatcher you call with is profitable except for the very bottom combination, which would be indifferent
It depends on how much you bet and how the board runs out, obviously. I wouldn't be at all surprised if there were some reasonable sizings/run-outs where getting 3 streets of value was possible at equilibrium. That doesn't necessarily mean betting will have a higher EV than checking at some point though.
The reason I'm so supremely confident is that I've spent some time studying the best HU limit bots, which are extremely close to playing GTO now, and after calling from the BB they c/f the flop often enough that c-betting is immediately profitable with ATC for the SB on pretty much every single flop. The larger bet size + larger ex-showdown equity advantage of being in position in NL makes it pretty much a lock imo.
I don't assume they don't, because they do.
Say you have two type of bluffcatchers, with and without card removal effects ("with" meaning beneficial for bluffcatching). If your range is tighter and you only need to call the bluffcatchers with card removal vs a given betsize, equilibrium strategy for BTN is to make those bluffcatchers indifferent. If your range is wider and you need to call the bluffcatchers without card removal, equilibrium strategy for BTN is to make those indifferent, and the good bluffcatchers just become slightly profitable calls.
There can be no such thing as an unprofitable river call when playing vs GTO, because you could just exploit it by folding
*edit* in theory it should be possible that every single bluffcatcher you call with is profitable except for the very bottom combination, which would be indifferent
Say you have two type of bluffcatchers, with and without card removal effects ("with" meaning beneficial for bluffcatching). If your range is tighter and you only need to call the bluffcatchers with card removal vs a given betsize, equilibrium strategy for BTN is to make those bluffcatchers indifferent. If your range is wider and you need to call the bluffcatchers without card removal, equilibrium strategy for BTN is to make those indifferent, and the good bluffcatchers just become slightly profitable calls.
There can be no such thing as an unprofitable river call when playing vs GTO, because you could just exploit it by folding
*edit* in theory it should be possible that every single bluffcatcher you call with is profitable except for the very bottom combination, which would be indifferent
Yeah ok, the point is that we both assume "for this runout and this sizing, hand xx is the bottom hand that goes bet/bet/bet in GTO". What we assume this hand to be decides our flop calling (and thus also continueing) range. I don't have a solid method to decide what hand XX is, do you?
Before Spladle started posting this stuff I was thoroughly convinced trying to approximate an actual NE on every hand was a waste of time as a practical means to learn poker but I've been slowly finding myself increasingly convinced post by post that it isn't. That it may in fact be practical and may even be superior.
I used an 48% 2barrel fwiw.
Probably for similar reasons as to why a strategy that only calls pre-flop would do worse than one that 3-bets.
The reason I'm so supremely confident is that I've spent some time studying the best HU limit bots, which are extremely close to playing GTO now, and after calling from the BB they c/f the flop often enough that c-betting is immediately profitable with ATC for the SB on pretty much every single flop. The larger bet size + larger ex-showdown equity advantage of being in position in NL makes it pretty much a lock imo.
What does the SB cbet strategy looks like (in limit) on a board like this then? How about say A82r?
You realize that the edit is in conflict with the rest of what you've said here, right?
As it happens, I'm almost positive that every single bluff-catcher you call with is profitable, only sometimes excepting the very bottom combination(s), which would be indifferent.
The claim that there's no such thing as an unprofitable river call against the Nash equilibrium seems insane and obviously wrong to me.
Not in pots with very high SPRs, no.
By took a look what do you mean? I think we should lay out methodology here.
Before Spladle started posting this stuff I was thoroughly convinced trying to approximate an actual NE on every hand was a waste of time as a practical means to learn poker but I've been slowly finding myself increasingly convinced post by post that it isn't. That it may in fact be practical and may even be superior.
Before Spladle started posting this stuff I was thoroughly convinced trying to approximate an actual NE on every hand was a waste of time as a practical means to learn poker but I've been slowly finding myself increasingly convinced post by post that it isn't. That it may in fact be practical and may even be superior.
Would like to hear more input on how you guys do this as well.
Of course. I used 75%. That explains our disparate results.
Meh, this feels like an empty one liner to me more than a good point. Could you elaborate on why you feel card removal makes check-raising 8x bad?
I don't think the precise strategies are very useful information due to the differences between limit and NL, but the BB c/f both far more often than necessary for the SB to show a profit betting any two, and yet the SB doesn't bet any two.
If the worst bluff-catcher you call with (where "worst" does not refer to absolute hand strength but instead to equity against the betting range, which is largely determined by card removal effects) is indifferent between calling and folding, then calling with anything "worse" than that will be unprofitable.
I could be a nit and point out that calling with the nut low will always be unprofitable because you'll still lose against "bluffs," but even if we limit ourselves to talking about legitimate bluff-catchers (i.e. hands that beat all bluffs), some will be profitable (or indifferent) and others will be unprofitable due to card removal effects.
An example from PLO may help to illustrate this. Consider the case of facing a bet on the river after check-calling the flop and turn on a board of Q7238. Obviously, QQ54, QQ54, 6543, 6543, and 6543 each beat all of villain's bluffs and none of his value bets - that is, they are all pure bluff-catchers. However, against the Nash equilibrium, I am supremely confident that calling with the first hand (QQ54) is unprofitable and calling with the last hand (6543) is quite profitable, while some other hand with between one and three hearts may be indifferent. See why?
I don't think so, no.
That's what I mean by unprofitable too.
Right, the possibility of raises messes everything up. Poker would be so much easier to solve if we could just get rid of that option. =P
If the worst bluff-catcher you call with (where "worst" does not refer to absolute hand strength but instead to equity against the betting range, which is largely determined by card removal effects) is indifferent between calling and folding, then calling with anything "worse" than that will be unprofitable.
I could be a nit and point out that calling with the nut low will always be unprofitable because you'll still lose against "bluffs," but even if we limit ourselves to talking about legitimate bluff-catchers (i.e. hands that beat all bluffs), some will be profitable (or indifferent) and others will be unprofitable due to card removal effects.
An example from PLO may help to illustrate this. Consider the case of facing a bet on the river after check-calling the flop and turn on a board of Q7238. Obviously, QQ54, QQ54, 6543, 6543, and 6543 each beat all of villain's bluffs and none of his value bets - that is, they are all pure bluff-catchers. However, against the Nash equilibrium, I am supremely confident that calling with the first hand (QQ54) is unprofitable and calling with the last hand (6543) is quite profitable, while some other hand with between one and three hearts may be indifferent. See why?
I don't think so, no.
That's what I mean by unprofitable too.
Right, the possibility of raises messes everything up. Poker would be so much easier to solve if we could just get rid of that option. =P
Apologies for slight derail but it's interesting because I've always believed, in fact thought it somewhat obvious that GTO calling strat out of the BB in HULHE called for >=30% flop folding. You regularly see people calling >80% which given positional disadvantage and rio even in HULHE is pretty bad. You just can't do that without making significant errors vs someone who adjusts.
Apologies for slight derail but it's interesting because I've always believed, in fact thought it somewhat obvious that GTO calling strat out of the BB in HULHE called for >=30% flop folding. You regularly see people calling >80% which given positional disadvantage and rio even in HULHE is pretty bad. You just can't do that without making significant errors vs someone who adjusts.
Averaged over all flops, it does appear that the BB folds to flop c-bets more than 30% of the time at equilibrium.
Feedback is used for internal purposes. LEARN MORE