dlayton66

"The thought experiment here is straightforward: if I can 3-bet any hand at least 0 EV against your opening range, I should."

no... you would never 3bet a hand for which callEV exceeds 3betEV, nor would you 3bet every hand for which callEV=3betEV>0 as there is no clear benefit to deviating from equilibrium in this way and counter-exploitation would be trivial.

"Any hand worse than 76s, then is -EV to 3-bet."

I don't think that's right. I can't think of a theoretical reason why that would have to be true, and in all the sims I've seen BB's range includes plenty of weak 3bet bluff combos which are indifferent between call/3bet, and even often between call/fold/3bet. I haven't done enough bvb pf work to say that sort of range construction is universal, but it's certainly not uncommon.

at the beginning of the video I don't follow your bit about the BB's 3betting range not being able to expand to more raggy bluffs when SB opens wider. why would you "making them indifferent" disincentivize BB from 3betting them? can you clarify what you mean by that? every 3bet hand that's not a pure value 3bet is going to be indifferent between either call/3b or call/fold/3b. I don't see why that means BB shouldn't 3bet them.

these days there is no such unspoken agreement between strong players. the reason most people don't often donk is because the PF caller is at both a range and positional disadvantage and thus wants to keep the pot as small as possible with a large chunk of his range (betting out with a ton of hands like someone else said is the shittiest theoretical way to realize equity). strategies also become exponentially more complex to analyze & difficult to implement accurately the earlier you split your ranges, so for simplicity's sake there is merit to playing a pseudo-equilibrium and never donking in a spot where equilibrium might see you donk some small % of the time.

if the PF raiser is OOP and faces a positional but not a range disadvantage, he still needs to check a decent amount on most textures, and should certainly not fire away with, say, all of his Ax on an A hi board as you indicated. that would be very bad. additionally, sometimes his range advantage becomes greatly mitigated by the flop texture and he should be checking even more still.

all the exploitive reasons for donking seem too opponent-specific to be generally useful. yes, some players will go bananas when you donk, but some players will play too tightly. you have to keep in mind that when you're making an inherently bad play you allow your opponent more leeway to make mistakes - even if he plays poorly vs your lead, your lead might be so poor in the first place that you still prefer to check. you also have to be very wary of your mental biases when you're trying to figure out stuff like that. the human mind is incredible at seeing patterns in noise. personally I don't see that there are any great population reads on this front.

ah poop. I misread the board.

still seems like a questionable bluffcatch. since we know claudico will make bets like this, we should have some flushes in our turn check-call range that are better calls. plus when he bets this large I think he'll occasionally be bluffing better, even when we hold the As. perhaps it's a small % bluffcatch, but it can't be that robust of a bluffcatcher.

the 2 blocker appears sexy, but realistically I don't think the bot is opening 52o, 52s plays earlier streets differently some % of the time, and many of the bot's potential bluff candidates have a 2 in them, so I don't really think this is as great of a bluffcatcher as it first appears. additionally with this action the bot can probably go down to some of the stronger 2pr for value, so the vast majority of cheet's bluffcatch range should be 2pr+. lastly if he calls too many hands like this the bot could counter by vbetting A7 for 20x pot, so that puts a strict limit on how much of cheet's bluffcatch range can include hands like this. wouldn't be surprising if A2o was a very small % call though.

he doesn't have to know we're folding 33 to exploit us for doing it

hi ben, would you mind sharing what the unlabeled numbers in your hud mean? thanks! good vid.

you are neglecting the fact that it's seiver. first, seiver is smart and likely exercises good BR management, i.e. he likely has only a small % of his action here. second, seiver is smart and plays well and will generally punish you (either by intent or accident, doesn't matter) for making bad plays. just not the dude you wanna be making insane explo folds against, though i think what you're saying prolly has some merit vs. other players

your criticism of equity distributions is silly.  most of the things you mentioned are captured in the visualization - the interaction of strong hands, med strength hands, weak hands, etc between the players, and though it doesn't directly represent a capture factor, just seeing how their ranges match up against one another is at least some insight into what that might be for each player (e.g. a player with a large advantage in the upper percentiles of their ranges will have a higher CF).  obviously the main shortcoming would be if one player had significantly more draws/nutty draws than the other player and would thus have a higher capture factor with a lot of his range that wasn't represented in the visualization, but that doesn't happen terribly often.  regardless, it's not even meant to be a complete analysis, just a starting point to very quickly (<5s) get a feel for how their ranges might interact with one another

barring an extreme equity disparity between ranges or one playing being extremely polarized, it's rare that GT doesn't involve some raises in almost every spot.  it's similar to the reason why we need to 3bet a certain % preflop to limit the EV of open raising - raising denies him equity, calling does not, so defense involving largely calling allows the EV of taking an aggressive action to be too high.

there's a fairly significant equity dropoff between Kx+ and the rest of their ranges, which as we can see occurs around the 20th percentile of hands for both players.  this board, then, is relatively drawy for a large part of both players' ranges, and neither player has a large range advantage, so it would actually be very shocking if equilibrium didn't involve both a xr range and a leading range for BB.

as to what hands we "want" to xr, there's vacuum incentive to xr almost everything.  Kx so we deny Ax equity & strengthen our raising range, 7x/2x so we deny overcards equity, sets for value/to strengthen our raising range, bluffs to force him to fold hands which have us beat.  all of this is level 1 analysis, but my point is that there are inherent benefits to aggression - we just have to balance those benefits with not bloating the pot with too weak of a range, not weakening the rest of our ranges too much, reopening the action, etc.

preflop is too complicated to know for sure; the truth is we're all mostly just guessing.  there is no reason button shouldn't be able to open ATC profitably.  there is no reason 72o should be a 0EV open.  there is no reason button shouldn't be able to cbet 100% of hands profitably (which is more or less what jungle said is the reason he was defending 43% in the video).  afaik the best we can do is to just guess at what preflop ranges seem reasonable using some pot captured estimations, hope we're in the general neighborhood of what's correct, and then play those ranges to be internally consistent

why do you have a problem having a 3bet range vs. UTG?  you haven't given any explanation for that

QJs >>>>>>>>> 54s in every preflop scenario imaginable.  54s is also an exceptionally loose 3bet vs. utg and probably means one of you is exploiting the other

also how the hell is a reg's utg open capped?

preflop seems like a clear mistake to me, i can't think of why you'd ever want this in your 3betting range.

not sure on flop, seems a bit loose... if you're 3betting J4s i'm guessing you 3bet a lot of depolarized stuff, in which case you probably have plenty of other hands that are way better for betting flop.  but still might be fine.

turn is a clear call with 18% odds.  you almost have that much equity even vs. super tight ranges, not to mention implied odds on the river, and of course he'll be bluffing or have some weird 77 sometimes in which case you easily have enough equity to call.  better fd seems pretty unlikely unless it's a pair+fd, but even then that's not many combos and they'd still flat sometimes too.

i'm not sure what he says in the video, but...

first, a 1/4 pot donk doesn't increase the amount pressure he's able to apply much at all.  i believe many turn and river solutions involve small donkbets like this even with an overall range disadvantage.  (although you'd probably never donk your whole range with a range disadvantage)

second, the 3 improves a portion of our range to the near nuts which allows us to defend well vs. someone raising and applying tons of pressure, so we don't mind increasing the pot size so much.  you seem to be glossing over this fact when you mention that we don't have strong top pairs in our range... we're protected by trips, and often wouldn't raise K7/K3/73 on the flop anyway.  that's why the 3 is a much better card to donk than the 2, although donking the 2 still makes some sense to me

third, the 3 is also good in the sense that we're probably calling some top x% of hands on the flop, and he's probably cbetting some range that is at least somewhat polarized, and the 3 misses a large portion of the somewhat polarized range which by default helps the top x% of hands guy

^these last two points are what make donking our whole range probably ok

take a multiway pot where one player folds before the river and the other two players reach showdown.  player A shows down with player B and wins a $13 pot.  player B lost $10 in the hand, which goes to his SD line.  player C lost $3 in the hand but folded earlier in the hand - because he didn't see showdown those losses go to his nonSD line.  the nonSD line for one player has decreased while the collective table's nonSD lines haven't increased at all.

no similar situation exists where one player can increase his SD winnings more than everyone else's SD losses.  when you consider that every pot not folded to the SB is actually multiway, this effect is actually very pronounced.  so in a rake-free environment with 6 players playing the same strategies, everyone will end up with a positive blue line and a negative red line that sum up to 0.

a subcase of that is where everyone is playing identically strong strategies.  if those strategies are GTO, anyone with a positive redline is in trouble.  however, since no one is playing GTO, all we can know from someone having a positive redline is that they're playing differently.

he might give up on this card, he might not.  impossible to say.  best we can do is call enough to make him indifferent to bluffing.  KJ no diamond is one of your better bluffcatchers since it doesn't block many of his bluffs, so i think this is almost certainly a call

looks good

i don't think you're at the bottom of your range.  you should have a few worse pairs, 97s/67s play like this very often, and 9Tss/9Tcc/9Tdd are also reasonable to play this way.  AQ is another hand that you might show up with as well.

even if you didn't show up with those hands i don't it matters as someone with high SD numbers is neeeeeever folding Jx here in a million years, and might not even fold TT, so the range of hands you are targeting is basically nonexistent

unless you call unsuited KJ/QJ/JT pre or flat 2pr+ on the flop, your range here isn't really stronger than his.  he should have around maybe 6-8 combos of boats (AA always plays like this, AJ/J9 do some % of the time), while you have 6 combos of trips, which really isn't very many at all.  his range is less capped as well as stronger on average, so i don't think he will have that difficult of a time dealing with a lead... add to this the fact that your turn checking range becomes incredibly well-defined when you lead Jx/draws, he can basically play perfectly vs. your checking range, and if you are leading a balanced range you still have to check turn pretty often since you don't have many value hands.  pretty clear check imo

typically the middle card pairing is a very good spot to lead since middle pair is not a great cbet, but here i don't see enough of a range advantage for you unless you show up with a lot more offsuit Jx combos

outside of lefort's theory videos (which are elite membership) there isn't really a curriculum per se, so aside from the handful of concept videos it may be more of a crapshoot whether or not a video touches on some concept you're interested in learning.  the videos are also much more focused on online play than live, so i'd just make sure to keep in mind the differences you've noticed between live play and online play when you're watching (e.g. people 3bet far tighter live, run big bluffs less often, etc.  online FR games are a billion times tighter than live FR games.)

mostly i'd just try to find some instructor who you like who thinks logically and articulates well and just watch all of their videos.  you want to single out people who don't just say "this is the standard play" but explain WHY it's the standard play, which unfortunately is a bit rare.  i'd recommend starting w/ james hudson's vids as he seems to play quite well.  while galfond plays mostly PLO now his NL vids are still excellent, so i'd definitely check all of those out also

this is an extremely difficult question and it'll be hard to give you a precise answer without devoting an obscene amount of time to it, so here are just some general thoughts i have:

1.  your goal isn't to be balanced... it's to take the highest EV line with every hand.  so if your range has a lot of SD value and checking a hand down is some EV, and placing said hand in a bluffing range is slightly less EV, it doesn't matter if your opponent has an easy fold with his bluffcatchers when you bet.  you still take the SDV.

i think what this should lead to is you being more polarized with your value bets.  so for instance if a hand is good enough to bet 3 streets for value (in the sense that a MDF'ing opponent's range would typically have <50% vs your hand by the river), now we have a situation where he's not incentivized to hit MDF since we're so weighted towards value, so the EV of value betting decreases, making the relative EV of checking increase.  so long as taking the bottom hand out of our vbetting range doesn't hurt the EV of our bluffs, we might end up having to do that.

however, the value of betting a bluff in a perfectly balanced range in a large pot is HUGE since you win the pot immediately vs. all his bluffcatchers.  it's going to be pretty hard to beat that return with a middling SDV hand, especially hands like 22-66 in this example which don't have that much equity and don't have an easy way of realizing it.   if your range allows for you to bluff that type of hand, then it definitely seems best.

essentially the scenario i mentioned above (wanting to vbet more polarized because you lack bluffs) will, i think, happen somewhat infrequently because you'll almost always have enough hands that prefer bluffing in the shade of a strong hand to showing down in the harsh, um... desert sun.  probably happens most on the river when equities are static.

2.  a hand doesn't have to be good enough to bet all remaining streets to bet the flop and/or turn.  first of all you have equity to protect on the flop and turn, so if a hand is good enough for 2 streets betting the first 2 could be best.  (this board is much less dynamic than others so that is less important here)

i recall sauce talking in a video once about value betting a hand for one street on the turn for a sizing that looked polarized because he didn't want his range for that sizing to consist only of hands that were betting the river since his opponent should never raise vs that range.  having a flop betting range of {hands that bet 3 streets, bluffs} is somewhat analogous to that scenario as he has little incentive to raise at any point until the river.  when you put a 2 street hand into that range, it makes his turn decision much more difficult.  now, it's not obvious to me that "difficult for him" = higher EV for us, so maybe we shouldn't focus on that, but i think constructing your ranges in such a way that your nutted range never overlaps with your nonnutted range is pretty obviously bad assuming he's balanced.

well, i don't know if i'd say no dearth of bluffs, since 98 or A6 aren't generally slamdunk bluff hands here, but still enough that i think we should be calling more than 20% of our range

if we assume your range for getting to the river is AK, AJ, TT, 99, Q7o, J6o, and 25% of AT/KJ, you have Qx or better only 20% of the time, so just having Qx occasionally doesn't protect you from bluffs here.  i realize i've assigned you a very specific range, but it seems pretty reasonable to me & it also seems unlikely you actually do (or should!) show up with Qx or better at a higher frequency than that even if the exact makeup of your river range is different.

we may not need to reach MDF here but calling all our AJ still only gets us up to 40% defense.  if he's a 400nl reg he'll likely be defending enough preflop to have no dearth of bluffs, so AJ is certainly a call.

it's possible KJ is a better call as well since AJ blocks more hands that might call preflop and bluff twice (e.g. K6 is not in his range but A6 is), but then you have to contend with the fact that he's probably playing AJ like this a decent chunk of the time, so i don't know which is better.  i'd just opt for the stronger hand unless i could pin down card removal effects with more certainty.

with the parameters of the problem i can't find a hand you'd rather call than 3bet (to 6.5bb).  of course he will pick up on that pretty quickly, but if this is the first couple times this has happened i see no reason not to 3bet 100%.  after the first couple times you probably need to dial that down quite a bit and probably just call A2o/75s.  if he continues opening 92% to min you never fold anything, and 3bet probably around 30% being very wide for value.

i struggle coming up with the best hands to bluff 3bet when calling with everything is profitable.  i think you would want to mostly just 3bet polarized since the weakest hands have the most to gain from 3betting as calling them is lower EV than calling T6s, but then again 82o is not as good in a 3bet pot as T6s.  if he 4bets pretty small and frequently and you find you need to defend some of your trash to his 4bets then maybe start opting for a more linear range

yeah i was curious about this as well.  86s is well outside of a top 15% range

AJ hand seems fine.  the 3bet is a little questionable and depends on his specific tendencies, but the rest of the decisions seem automatic given that 3bet.  he is leggo instructor apotheosis fwiw

no offense but you rambled on way too long here, like how people used to never flat 3bets oop.  ofc we know it's ok to flat 3bets oop now.  that didn't leave you much time to talk about hands which is what the people want!

I don't see how you can say this without knowing their ranges.  like ben pointed out in the video, none of the plays were provably bad (except for possibly the 99 minbet), and the fact that they utilize so many different betsizes in what could be a balanced way opens up the possibility that they would crush any human playing now since that is something very few people do but is something that we know is good in theory.

of course I can't assert that the bots would beat a good (or even mediocre) player, but nothing I saw indicates to me that they necessarily wouldn't beat a worldclass player HU.  with how little information we have to go on we could probably only say these bots are anywhere from 1/2 hu reg level to world class, and neither end of that spectrum should be surprising

many of the plays that are considered standard now would have been considered terrible 5 years ago.  we have a tendency to label things that are unfamiliar to us as bad, but if we want to grow and learn we need to avoid that trap and be as open-minded as possible

one of the bots did half pot J7hh on 338hhQh on the turn, and earlier one of them went for a x/r with the 9h on a 4 heart board, which counts as a slowplay in my mind (since the rest of the hand was checked down the bot views that hand as quite nutty it seems).  have yet to see a minbet or 1/4p bet balanced, but it is hard to make strong hands so we don't have a great sample size on that yet.  balancing an individual betsize with a strong hand is such a basic concept though that I have to imagine they do this at least in some capacity

sorry but you are wrong.  he will have essentially every combo of ATo-AQo here in addition to many A7cc type hands and perhaps even weaker Ax than that since he was described as sticky