postwar18's avatar


489 points

11:48: I agree that IP is unlikely to be mirroring solver strategy pre-flop, especially with regards to 3-bet-calling mid pocket pairs. Removing these hands from IP range, you derive two fairly similar ranges, both consisting of many broadway-type hands and high pocket pairs. If you, as OOP, use a half-pot C-bet size, you're relying on IP folding some non-intuitive hands including AJxx and AQ immediately on the flop. It seems doubtful this will happen in practice. Therefore, your instinct--to give up with this hand--seems reasonable. Of course, it's important that if you are going to bet some of these hands, you use clubs to bluff: they have the most equity but can't continue against a bet.

April 26, 2024 | 3:26 p.m.

25:00: I understand the instinct to raise the turn given what you said about OOP leading strategy: that many players do so with hands that look right, rather than based on an understanding of the two ranges, and which class of hands, by betting, they're trying to target (along with a corresponding appreciation of which parts of their range, by checking, they're trying to protect). Against this sort of 'merged' strategy, might not one be more incentivized to bet the flop with hands like yours? Presumably, opponents such as this one--who lead overly merged on the turn--are not going to be over-check raising flop. Your hand might not be the sort that loves being check-raised on the flop, given that it has to call, but isn't happy about contending with future aggression against a polarized range, even when IP. However, by checking, it seems you might be affording OOP some prerogative to determine how much money goes in the pot with mid strength hands and draws. Even in this spot wherein, as you acknowledge, OOP plays turn and river poorly, he still might have been made to pay more. For instance, in this exact spot, after betting the flop, you might comfortably go large again on the turn, expecting more passivity, and potentially even some over-calls with marginal made hands or hands with equity that you largely dominate. On the river, even, you might have potentially bet for exploitative reasons. Thereby, after betting the flop, you might have had more strategic options on later streets to get more money against a large class of hands I have to believe OOP would both over-call and, on the whole, play too passively.

April 26, 2024 | 12:17 a.m.

20:00 table 2: Finding bet-folds seems tough. What hands would you use? I'm not sure this SPR is one very conducive towards bet-folding against a solver. However, facing a human, I can see you garnering more folds than you would against a more equity-driven computer. I imagine this might work particularly well against certain classes of hands--AA, for instance. Are marginal made hands, presumably without much backup draw equity, those which you are seeking to challenge by finding bet-folds? Are there certain classes of hands in your range you might feel more comfortable bet-folding against a conservative human opponent than against an equity-driven solver?

April 25, 2024 | 11:50 p.m.

14:41 HU table: I think solver will have a few more flop slow plays and turn raises that one might face from humans in this turn spot. Thereby, I think that the risk one runs by potting a draw like yours is minimized. Similarly, your hand becomes very difficult to play as a check against an IP polarized turn sizing. Therefore, bet seems good, especially if you're discerning regarding river bluffs, which it appears, based on your river give up, you are. I would think that in spots like these, facing a condescend and largely transparent IP range, that the OOP player would like to take the initiative and dictate terms.

April 25, 2024 | 5:15 p.m.

5:00 table 2: I assume that, in addition to the pre-flop variables you mentioned, i.e. small blind having more suited aces, that, by the river, there is a large asymmetry in ranges whereby IP has many more straights and many fewer flushes. I imagine that with the nut straight, in particular, OOP is rarely going to be slow-playing the turn. Similarly, OOP is probably not going to use too many of his flush draw combos, up to and including the nut flush draws, as turn semi-bluff raises. Conversely, I imagine that IP is incentivized not to bet an exceptionally large proportion of his mid-strength flush draw combos on the turn, fearing that he might be compelled to bet-fold, and fail to realize their equity. What's more difficult to find are OOP river bluffs, particularly for pot. Presumably virtually any hand with the Ah blocker that is less than a straight is playing this way? If so, how often is IP compelled to call with a straight?

April 25, 2024 | 4:44 p.m.

What strikes me most significantly about the spot is the impulsivity OOP displays: not only does he quickly decide not to bet the turn, which game theory overwhelmingly dictates he should, he afterwards quickly checks on the river after hitting a card that overwhelmingly changes what he's trying to achieve. Whereas on the turn, based on my conjecture, his misbegotten instinct is to attack an unexpected, small bet-size, without much regard for the the future playability of his hand; on the river his immediate instinct is to stick-it-to-you or go for a check-raise, in a spot he has many other options for trying to get as much money as possible, i.e. overbet, bet-3B, etc. It's hands like these, in my opinion, that allow one to derive a read against these type of opponents that can then be used more globally, in many future spots: it's what inspired me to characterize his street-by-street tendencies based on the evidence of a single hand.

Nevertheless, one must make these decisions hand-by-hand as one finds them. My overall conclusion is that he plays impulsively, and operates, street-by-street, with an inclination towards check-raising for value against opponents he feels bet too often IP; this informs my conjecture that in these sorts of spots he likely is inclined towards value relative to bluffs; and dictates that one should likely over-fold this particular node. Still, one does not put one's eggs solely into one basket, as it were, and impute that he only has value in this spot. Rather, you may rely on the dictates of game theory, as you did, to call certain hands that strongly block his primary value hands, while at the same time over-folding other hands that against a more balanced range, you might obligated to call at some frequency.

April 24, 2024 | 4:45 p.m.

I'm inclined to think villain isn't necessarily thinking in the terms you impute he's employing in this spot; that is, you think in terms of equities and hand classes, whereas I believe him to be responding more reactively and with less distinct reasoning. You're right that from a GTO perspective, he should almost always be betting his hand on the turn: there's overwhelming impetus that he get value immediately: 55 is strong enough against your fairly weak flop check back range, that it can bet every time for pot, comfortably; plus, it is strongly incentivized to get immediate protection against overcards that will fold, i.e. A7, KQ, K9, etc. Conversely, against the size that he expects you to bet IP, i.e. 75%, 55 has a lot more trouble playing as a value check-raise; it might check-call, but that isn't intuitive, and doesn't really accomplish anything. Therefore, I think by mindlessly checking 55, OOP reveals that he probably doesn't have strong theoretical reasons for doing so and is playing by feel. Your unexpected turn sizing puts him in a spot wherein his instincts change and he goes aggressive, feeling you're very weak and that he hasn't been able to put enough money in with his hand, so far. Then he gets a gin river, and he goes from wanting to put more money with his hand on the turn, deny some equity, and reevaluate on the river, into an instinctual passive mode wherein he expects you to be the sort of player who bets too often when checked to, especially using small sizings, as a matter of course. I don't think he expects you to bet many ace-highs, or hands as strong as 3x, as bluffs, whether you should in theory or not. Similarly, I don't know if he expects you to have more Tx that you should after checking the flop, or thinks of the problem precisely in those terms. Nevertheless, it seems like he's playing reactively, and that his experience has been to expect a lot of bets in spots wherein, in theory, you as IP shouldn't necessarily have that many.

That being said, I think you speculated that he wouldn't necessarily be thinking to play straights this way, and I agree. Not only would he not want to go for stacks with straights on the river, it doesn't seem like he'd be the type to play his straight draws this way on the turn: expecting a 75% sizing on the turn, he'd be in a pretty rough spot, especially with 67. Moreover, he doesn't seem like the sort of player who would be inclined to play this way with 65 on the turn, regardless of what theory says. Therefore, I think you still reach the appropriate conclusion: that he would only make this play with boats; and, you, in turn, make the best decision: calling, given that you have the Th in your hand. Still, given that I think he'll have less 65 that he might otherwise in theory, I don't really know from where he might be deriving his bluffs in this line.

April 21, 2024 | 3:29 p.m.

It must take tremendous courage for unsympathetic dolts to mindlessly and hypocrtically lend their support to those reliant on their position, rather than their intellect or capacity for thoughtful repoonse, to persecute the defiant. Kudos to you if such ineloquent and unctious blather is all it takes to satisfy your pride and sense of solidarity.

April 16, 2024 | 3:43 p.m.

Comment | postwar18 commented on 1 Hour of $500 Zoom

29:07 table 1: I think that, in theory, you're supposed to call Tx rather than 3x vs this sizing, especially if you're correctly polarizing on the turn as IP. However, given the hand OOP showed up with, it seems OOP is likely seriously over-bluffing in this spot; at the very least, OOP should be limiting himself to his ss and cc trash if he's betting a size that reps a Q. It appears KT and AT are also value bets for this size, although I doubt OOP is going to as readily bet them for value as he is to pull bluffs from parts of his range that can't support it. Therefore, I like your call.

April 15, 2024 | 11:06 p.m.

What a load of inarticulate nonsense. I am annoyed because I think you’re a petty reactionary less concerned with responding to the merits of the criticism than that it was made in the first place. Furthermore, simply claiming not to be involved in self-justification doesn’t make it so; it’s a vulgar form of sententious posturing.
In this case, you were value-betting—unsuccessfully. Show me some instances in which you save money by bluffing for the smaller size, rather than the larger, in the belief that opponent will over-fold his ‘air,’ and then you can talk: I’ve seen some of your other videos and I can’t recall you ever, in this line, successfully bluffing for the smaller size knowing that your opponent will over-fold a certain class of hands, regardless of the size of your bet; nor have you proved that IP might overbluff facing a small bet, quite the opposite, in fact. IP justified his decision according to his own misunderstanding of GTO; his fold goes no way towards validating your bet, especially since I’ve already said, but apparently must repeat, you had the nuts. If you’re so confident that IP will over fold his ‘air’ (the way you class hands that will fold to either small bet or large bet), maybe you should have checked, and tried to maximize your EV against overbluffs. As matters stand, you certainly didn’t maximize your EV in this hand.

April 13, 2024 | 10:02 p.m.

Comment | postwar18 commented on 1 Hour of $500 Zoom

6:55 table 3: I understand the principle of wanting to call with JJ that block the Qx suited. Interestingly, according to solver, apart from actually being inclined to bet this combo more than others on the flop, it also wants to jam turn, forcing a lot of flush draws with a lot of equity to fold. On the river, it prefers calling JsJc, more than any other JJ combo, facing jam, because IP is expected to give up with cc and ss bluffs and triple barrel with hh and dd combos. According to theory, then, you're more concerned about playing back against the manner in which a GTO opponent plays their flush draws and zero-equity triple barrels than you are about making call-downs according to the frequency with which they'll have Qx. In practice, I don't think there are many IP players who triple barrel JdTd or 9dTd in this spot, however; whether that means that they'll be overbluffing with missed ss and cc combos, I don't know: presumably not. Altogether, seems like a spot where you can overfold bluff-catchers on the river, particularly when you have more of them than might be expected, according to GTO.

April 13, 2024 | 7:37 p.m.

Finding the bluffs in the block/pot/block line is a lot more difficult than finding the value bets, even if you know from whence they might be drawn, itself no easy task: the solver mixes a large variety of hands include pairs and high card hands that might also be used to bluff for the larger sizing: I seriously doubt most humans' abilities to execute a strategy at all closely aligned with theory. I understand that you're inordinately gifted; but even you, I'd presume, would have trouble, in practice, in mixing your overbet and block bluffs in a manner so as to best target those Qx or Kx hands for folds. I think anyone, knowing they were trying to target those hands, would be inclined to bluff larger more frequently than the strategy employed by the solver, trying to ensure a fold, rather than attempting, in an exceptionally complex and sophisticated manner, to push those hands towards indifference against multiple sizings. As for those A7-A9 hands, I don't see what that has to do with anything. It means you'll have less value, which, I'd assume, would mean you'd also be allowed fewer bluffs; that argues against carving out a smaller-sized bluffing range. Although I might be 'mental,' I wasn't arguing that IP wouldn't raise AJ against block-pot-block. Rather, I don't think he'd go all in with it, something you both acknowledged in the video. Perhaps, he might almost always call against the 3B shove, but that's hardly of any consequence. According to the sim you used, facing a GTO villain, you'd be getting stacks in by the river anyway.
Finally, as far as AJ goes, the same principle applies: villain won't have slow played enough of it, in practice, by the river. Even weaker players, in my experience, will understand implicitly how capped their ranges are by the river, and will refrain from making as many bluff raises as they'd be allowed were they playing a GTO strategy: blocking with the nuts is a strategy that suffers significantly when IP isn't bluff-raising as much as one, according to a GTO sim, would expect.

April 13, 2024 | 7:01 p.m.

last hand: I think BVB OOP block, pot, block is a line wherein population lacks bluffs, especially on boards in which the nuts doesn't change.* The only circumstance in which this wouldn't be the case is a situation in which OOP expects too many bluff raises on the part of IP; in this hand, that hardly seems to apply, however. If anything, it's pretty unlikely IP slowplays as much KQ as is expected by the solver. Therefore, not considering raising the river, whether it be with this hand or some other Q blocker that in theory does more raising, seems fine. Conversely, blocking the river without a reasonable expectation of facing bluff raises, or for that matter large, thin value raises (e.g. AJ) seems bad. Furthermore, overbetting the river pushes Ax hands, which you unblock, towards indifference. Depending on the dynamic you've established with the IP player--for example, if you've previously challenged their lack of flop slowplays by bluffing large on turns and rivers aggressively --you might get them to overcall river with Ax facing overbet.

*The sorts of hands one uses as a bluff and the way one goes about mixing those hands is also excessively complex

April 12, 2024 | 1:49 a.m.

28:00: I don't like double barrell turn sizing or combo. Your size does not target his Ax hands for turn folds, which I think is desirable. Moreover, the sizing puts you, IP, in some tought river spots on relative blanks, e.g. 2c, wherein you don't really know what to do with hands like AQ: they want value but they don't want to go all in (as opposed to overbetting AQ on turn and checking river, thereby making more money overall)*. Finally, QsJs is particularly bad in that unblocks all his Ax suited flop calls. Overall, this means you're not going to get nearly as many folds as you would like with this hand.

*This is particularly the case against OOP's apparent suboptimal strategy wherein he's failing to raise his sets on the flop or turn. This allows you to really punish his range with larger turn bets. Against them he's going to be calling an inrodinate amount of weaker Ax and not put you in a difficult spot by raising and playing for stacks.

April 11, 2024 | 7:48 p.m.

27:18 I agree with the fold generally. I think you're also on to something regarding calling with the low kings. It seems, according to the preformed sim upon which you relied, that even given a strong GTO opponent, one calls more of the low kings than the higher ones, assuming IP is composing his turn bluffs mostly from the broadway, turned gutters, that he didn't bet on flop. The only reason OOP might still include mix some folds with the lower kings and mix some calls with the higher ones is that in addition to those sorts of turned broadway gutters, IP might turn some low pairs into overbet bluffs on the turn. I think in practice, this is never the case. Some strong players might overbet bluff low-pairs on the turn. Nonetheless, virtually no one seems to mix checks with their low pairs on the flop and then decide to use them as overbluffs on the turn.

April 11, 2024 | 7:23 p.m.

26:00: OOP almost never bluffs SS or CC combos on the turn, therefore, if he is betting those hands he has an inordinate amout of value to bluff combos. Conversely, IP bluffs relatively more hh and dd hands hands, e.g. QdJd or QhJh. Therefore, with a range of indifferent Kx hands--not just KQ, but other suited Kx hands as well (the Kx folds facing overbet vary according to the parameters of the sim, but the principle doesn't)--one is generally more likely to call non DD combos facing overbets.

April 11, 2024 | 7:14 p.m.

25:27 Table 1: not a big fan of flop Cbet. Seems like the sort of hand wherein you have just enough equity, between the wrap and the weak backdoor flush draw, that you'd rather realize it than bet-fold. I'd guess one would prefer hands with one diamond blocker, that has no equity but blocks raises and can potentially bluff diamond runouts, or hands with, for instance, a jack and a pair lower than 99 that blocks value, but doesn't stand to lose equity at all by folding much by folding. Another consideration would be what sort of hands you're looking to fold OOP off of by betting 1/2 pot. Presumably you're targeting some of his weakest gutshots without much going for them or his weakest Jx. It seems you could probably get those hands to fold anyway, on later streets, if none of your draws materialize and you're left with Q-high. The flip-side of that is that OOP might himself try to bluff some of the weak-gutshot type hands on later streets that would have simply folded to flop bets. However, it seems, assuming flush draw doesn't come in you can still continue against most turn bets against most turn cards with this hand, facing a bet; then, if OOP bets turn and checks river, you can bluff (assuming you continue to miss) knowing that OOP bet turn check river is usually a fairly weak, give-up range.

April 11, 2024 | 6:37 p.m.

24:00: IP bets an absurd sizing (from a mathematical perspective) on the flop: as we see, it's almost never used according to GTO. The very little the solver does use this middling size seems consonant with the exact strength of particular middle-strength hands, namely KTo and QTo. Against these hands, on the river, your 1/4 block loses a tiny bit, but not much at all. Conversely, I think you have a lot to potentially gain by blocking against A8 type hands, which IP will have in his range much more frequently than the solver (relative to QTo and KTo). Assuming IP never turns hands like A8 into value raises, which is likely the case, your block is good.

April 9, 2024 | 4:02 p.m.

30:00: I think the most important consideration, by far, in these spots wherein you''re considerating a semi-bluff jam, is how often OOP is bet-folding the hands that dominate yours; for instance, how frequently he's bet folding AK, AQ, KQ. At GTO frequencies, your hand is already a jam. Against opponents who are going to be betting an inordinate amount of AK, AQ, KQ, it becomes enormously profitable.

April 9, 2024 | 3:47 p.m.

20:00 Good river check. It appears that, because IP value-jams Qx when checked to, by putting AA in the x/c line, you go some distance towards protecting your range against overbluffs. As for the turn, I think it prefers the check without the diamond to put the IP Add hands in a terrible spot after they bet when checked to, which they will do at a very high frequency: these hands are then forced to call off versus a hand--AA-- against which they have very poor equity. Similarly, AA no diamond can potentially induce a small number of bet-folds against other dd hands, denying them a decent amount of equity immediately on the turn.

April 8, 2024 | 6:42 p.m.

17:20: I agree with your assessment that villian will bet too much one pair for value on the river relative to GTO as well as include too many corresponding bluffs, incentivizing you to x/r less frequently on the turn and lead less on the river than GTO would suggest (in addition to the fact that playing a lot of turn x/r and river lead makes your overall strategy more complex and thereby more difficult to play). This overly-thin IP value betting strategy also has implications on your OOP flop strategy. Facing villians who consistently bet too thin over multiple steets on blank runouts incenvtives OOP to slowplay more hands on the flop, I would think. You become less concerned about targeting certain IP hands for indifference on flop and turn with x/r (as well as fighting for equity with some of your draws like 9T) and more preoccupied about constructing an inordinately strong check-call down range against IP who will put too much money into the pot with marginal hands and bluffs.

April 8, 2024 | 5:04 p.m.

To qualify: I usually almost always overfold AQo in these spots by default, even when I have a vague feeling that I might be overflolding facing a GTO range. I typically have some implicit understanding that I, personally, don't always play these complex spots competently OOP; and I don't want to put myself in a situation wherein I have to defend a weak and/or capped range on later streets. Even somewhat weaker players can usually, as IP, figure out ways to put a lot of pressure on capped ranges on turn and river. Moreover, it might be argued, that IP overbet flop ranges, on boards such as these, are typically composed in such a way that calling the AQo hands are losing immediate EV. This happens, for instance, when IP fails to find a lot of those marginal QTo, KTo, KQo type overbet bluffs and instead bets a lot for protection with hands like 99 or TT, or as a semi-bluff.

April 8, 2024 | 5:03 a.m.

Yeah, I suppose you're right: it does cost a lot to of immediate equity to overfold AQo, KQo, QxTx hands OOP versus overbet. I was considering the implications of OOP having to call some of these hands, and because of a suboptimal x/r strategy, facing extremely negative EV decisions with them on later streets. In other words, I think that villains who don't understand what sorts of overall strategy they should play against overbet are bound to lose with these hands, regardless of whether they overfold flop or later streets.
Against overbet, it appears, according to the solver, OOP wants to be playing a small, almost click-back strategy to battle those sorts of KTo, QTo, even QxTx backdoor hands that gain so much against AQo type flop folds. I think against the vast majority of villains, this is hardly ever the case. If villain is not playing this click back sort of strategy; and, like you said, is playing a more polarized x/r strategy, your KTo and QTo hands do really well betting the large size, both in theory and especially from an exploitative perspective. I was surprised to see that facing 125% on flop, OOP, when limited to the option of raising pot, hardly ever raises 88 or 66. It seems that if he isn't playing a strategy wherein he's trying to make those sorts of low-equity overbet bluffs from IP indifferent on the flop, that he has to keep his range unusually strong so that he can better contend with IP aggression on later streets. Playing a polarized x/r strategy that includes too many of your strongest hands, puts OOP in a position wherein his flop calling range is quite weak and capped: it means that against a strong IP player, he's going to be unable to meet defense thresholds on future streets. So, while it is certainly negative EV to overfold those mixed defends immediately on the flop, it seems unlikely you'll lose much more than you would by calling them at reasonably correct frequencies, but also playing a suboptimal x/r strategy, or not realizing that your weak calls need to be protected by slowplays. Invariably, it'll leave your flop check call range overly weak and capped and expose you to a lot of EV loss against IP turn and river bluffs.

April 8, 2024 | 4:02 a.m.

16:54 I agree this makes most sense as a flop Cbet, given turn playability. I'd assume you're potting the turn with this hand most of the time primarily to target folds from IP with some of his bare Kx, and to still get called by pairs and draws your hand dominates? As played, I'd agree that IP has an inordinate number of hands exactly like the one he showed up with --K52--and semibluffs. Still, by the river, IP should have the nuts more frequently, which makes you want to range check; resultantly, you'll end up getting more check backs than you'd like with middle strength hands like these.

April 7, 2024 | 11:49 p.m.

43:56: Interesting observation--the difference between OOP C-betting strategy BB vs CO and OOP SB vs CO--and the latter being a range bet. I looked into BB Cbet and CO counter-strategy and remarked that what I see in practice, is in no way similar to what one sees in GTO. BB is expected, in theory, to mix small bets and include a bunch of check-raises against a relatively depolarized and widely mixed CO strategy. I think, in practice, BB checks a decent amount of his moderate strength hands, and doesn't necessarily play much of a x/r strategy. Conversely, CO will often polarize facing check, and BB will play a passive call down or fold strategy (although solver much prefers battling via check-raises). I would guess that SB would misapply this principle to SB vs CO spots as well, which would make me assume that he has either a nothing hand, like the hearts you mentioned, or a moderate strength hand way more often than he would AK or the 99 that actually showed up in his hand.* Therefore, I think assignming him an inordinate amount of HH bluffs in this unforeseen node makes this a good call down.

*As you menionted, 99 is a particularly absurd hand to play this way. Using a more conventional small range C-bet on the flop, bet turn and jam river, he would still, more than 2/3 of the time, get all the money from your hand AJdd by the river.

April 7, 2024 | 11:15 p.m.

28:42: Does IP do very much bet-folding with stronger two pairs?

April 6, 2024 | 11:25 p.m.

19:55: Gets really interesting to start check-raising this hand, A9, if IP overfolds to large turn x/r size. In situations in which opponent pure folds hands like QJ (as opposed to mix-fold), you can even, it seems, x/r hands beyond A9, like Q3, targeting better Qx to fold on the turn. On the river, you could then pure check assuming IP, in addition to folding too much Qx on the turn, will call a disproportionate number of missed draws: this makes it easy for him to over-bluff the river, especially if he doesn't have a tight grasp on his value thresholds.

April 6, 2024 | 10:18 p.m.

15:36: is 1/3 the only size you're giving OOP to continue betting on turn?

April 6, 2024 | 7:54 p.m.

Another interesting facet of the hand, as illustrated by your final graph in which BT shoves when checked to, are that certain hands, a bare 7 or 8, for example, fold to an inordinate degree relative to how frequently they shove OOP. I think doing a deeper dive into the nature of what sorts of hands prefer targeting IP folds by betting vs what which are too weak to call vs IP pots might prove enlightening.

April 6, 2024 | 7:39 p.m.

9:00: I appreciate, very much, the concept of showing what, by shoving, you're targeting for indifference. I think it'd be improved in this particular graph by forgetting about what calls vs what shoves, which seems of little consequence, relative to the frequencies associated with the various hand classes delineated in the graph. For example, what percent of IP 4-bet air hands, most of which are folding, constitute within the overall IP range? Moreover, how much equity do these hands, which as OOP you're targeting for indifference, have against OOP range?

April 6, 2024 | 7:34 p.m.

Load more uses cookies to give you the best experience. Learn more about our Cookie Policy